I am a faculty member in the Department of Mathematics at Saintgits College of Engineering with over 20 years of experience teaching pure, computational, and applied mathematics across science, engineering, and technology programs. Alongside my faculty role, I am currently a Research Scholar at the Amrita School of Artificial Intelligence, where my work bridges deep mathematical theory and cutting-edge machine learning paradigms.
In addition to my academic roles, I serve as a Data Analytics consultant for engineering and social sciences, and as a National Resource Person for AI & ML training initiatives led by prominent institutes like IIT Madras and the Government of Kerala.
My research addresses the intersection of mathematical modeling and deep learning, with a specialized focus on medical image restoration. I work extensively on deblurring and reconstruction techniques for clinical imaging modalities (CT, MRI, X-ray) to bridge computational theory and practical clinical requirements.
Core Research Areas:
- Custom Deep Learning Architectures: Developing mathematical formulations for specialized network components, including the Spatially-Adaptive Gated Activation (SAGA) and the Augmented Differential L1 (ADL1) loss function.
- Clinical Imaging Optimization: Formulating physics-informed models and carefully evaluating architectural choices, such as the deployment of SiLU and ELU activation units, to enhance interpretability and accuracy in clinical contexts.
- Collaborative Medical AI: Partnering directly with clinical experts (including radiologists from the Amrita Institute of Medical Sciences) to ensure models meet rigorous diagnostic standards.
Recent Publications:
- An interpretable deep learning method for medical image deblurring and restoration DOI: 10.1016/j.health.2026.100468
- X-GAN: Physics-informed generative model for edge-preserved chest X-ray image denoising DOI: 10.1016/j.cmpbup.2026.100247
- EffNet-SVM: A Hybrid Model for Diabetic Retinopathy Classification Using Retinal Fundus Images DOI: 10.1109/ACCESS.2025.3566073
Software Release
saga-activation: A Spatially-Adaptive Gated Activation (SAGA) with Fused Triton Optimization. documentation
I am deeply invested in training the next generation of engineers and researchers. To date, I have led over 40 Faculty Development Programmes (FDPs), Short Term Training Programmes (STTPs), and technical workshops across India.
- National Resource Person: Selected as a trainer for the AI & ML program offered by the Government of Kerala with technical support from IIT Madras, leading extensive, multi-month certification programs in Reinforcement Learning, Statistical Methods, and Mathematical Foundations of ML.
- Institutional Mentorship: Serving as an Institutional Mentor for the Intel-Unnati program, bridging academia and industry standards.
- Curriculum Development: Actively involved in structuring academic frameworks, serving as a Board of Studies (BoS) member in Mathematics at Assumption College (2023β2025) and as an External Subject Expert (CS & Allied Programs) for APJ Abdul Kalam Technological University (KTU).
- Workshop Themes: My training sessions primarily cover Artificial Intelligence, Python and R for Data Analytics, Scientific Computing, and the mathematical foundations (Linear Algebra, Probability, Optimization) essential for Machine Learning.
- First Rank in B.Sc (Mathematics) and Fourth Rank in M.Sc (Mathematics), Mahatma Gandhi University.
- Scholarship Recipient for Post Graduation in Mathematics from the National Board for Higher Mathematics.
- Best Project Proposal Award from RACCS, St. Thomas College, Palai.
- IUCEE Certified teaching faculty through the International Engineering Educators Certification Program.
- Scientific Computing & Analytics: Python, R, SPSS, MATLAB
- Machine Learning & AI: Deep Learning, Reinforcement Learning, Generative Models, Interpretability
- Mathematical Applications: Computational Optimization, Statistical Modeling, Random Processes
-
Scientific Publishing & Documentation: Advanced
$\LaTeX{}$ ,Quarto(for reproducible research and dynamic reporting), customized templates, and automated reporting